FAMILIES OF ALGEBRAIC TEST EQUATIONS W. GAUTSCHI (I)
|
|
- Britney Clark
- 5 years ago
- Views:
Transcription
1 FAMILIES OF ALGEBRAIC TEST EQUATIONS W. GAUTSCHI (I) 1. Introduction. Test matrices have been in use for some time to scrutinize computer algorithms for solving linear algebraic systems and eigenvalue problems; see, e. g., Gregory and Karney [1969]. For the problem of finding roots of algebraic equations, the construction of appropriate test equations has been given less attention. Here we wish to propose two families of algebraic test equations, the first consisting of equations with predominantly complex roots, the second of equations with only real roots. To be useful for testing purposes, a test equation (of some fixed degree) should have the following characteristics: (1) All roots of the equation can be calculated directly (i. e., without recourse to a rootfinding algorithm). (2) The equation contains a parameter (or parameters) which can be used to control the numerical condition of the roots. By varying the parameter (s), the condition number of the worst-conditioned root can be made to range from relatively small to arbitrarily large values. (3) All coefficients of the equation are integer-valued. It may not be easy, in practice, to achieve all these characteristics, particularly the last one, if we are interested in relatively large degrees. Even when this is possible, the integer coefficients may become so large as to make exact representation in floating-point arithmetic impossible. Although equations which do not satisfy (3) are less desirable, they are still useful for testing purposes, Received ]une 6, ) Department of Computer Sciences, Purdue University, Lafayette, Indiana 47907, U.S.A..
2 384 W, GAUTSCHI: Families provided one takes properly into account the influence of rounding errors in the coefficients upon the results. Isolated examples of test equations, particularly ill-conditioned ones, have been known for a long time. Perhaps the best-known example, due to Wilkinson [1963], is the equation with roots at 1, 2... n. Some of these roots are relatively well-conditioned, while others are quite ill-conditioned, more so the larger the degree n. (The numerical condition of Wilkinson's equation is analyzed in Gautschi [1973]; see also Gautschi [1978, w 4]). The roots of unity lead to another interesting example if one removes half of them and retains only those on the half-circle (lenkins and Traub [1975]). Our first family of test equations, indeed, is a simple extension of this latter example. 2. A first family of test equations and their numerical condition. Given an algebraic equation of degree n, (2.1) p (z)=0, p (z)=z~+al z~-l+...+a,,_l z+a,, with complex coefficients ax, and a simple root ~ of (2.1), we adopt as condition number for ~ the quantity (Wilkinson [1963, p. 38 if], Gautschi [1973]) (2.2) cond ~-- z rail Ir ]p' ( )l " It measures the sensitivity of ~, in terms of relative errors, to small (relative) perturbations in all nonzero coefficients az. Our first family of test equations (with parameter a) is (2.3) p~ (z) =0, p, (z)= /-) [z-r (a)]=z"-~ z"-l+...+(-1)" ~,, 2~ r (co =e(~-')% O<~z_< -- n If ~z=21r/n, the ~ are the n-th roots of unity, thus p~ (z)=z ~- 1, and we are in the case of a well-conditioned equation, all roots having condition 1/n. For a=rr/(n--1), we get the roots of unity on a half-circle, which are all relatively ill-conditioned. As c~ $ 0, the condition deteriorates unboundedly,
3 o~ algebraic test equations 385 From Eq. (2.2) we find for the condition of the roots in (2.3), (2.4) cond (~= ~=', v=l,2... n. a,~, (Z 2 ~'-I H sin (v--2) -~- In view of the symmetry property cond~=cond~+l-~, v=l,2... n, it suffices to consider the condition numbers for v= 1, 2... n', where We show that In + 1] n'=[ 2 I" 27~ (2.5) cond ~1< cond ~2<...< cond ~, for 0<a< --. ll If n=2, there is nothing to prove. If n>2, we let zr~ denote the product in the denominator of (2.4) and observe that (2.6) ~v-1 sin (v-- 1) 2 sin (n--v+ 1) 2 v=2, 3... n'. Our assumption on a implies that both sine functions in the numerator and denominator are evaluated in the open interval (0, re), the former in fact in (0, re/2). The absolute value signs in (2.6) can thus be deleted, and an elementary calculation shows that all ratios in (2.6) are less than 1 precisely if ~Z ) 6~ /,/,. (2.7) tan ~t~- > tan(v--l)-~-, v=2,3... Since the tangent is monotone increasing on [0, z~/2), the inequality (2.7) for v = n" implies all others, and indeed holds true by virtue of,(o Itan n 9 if n is even, a _. 4 2 tan (n'-- 1) -- tan n 4 4 if n is odd. It follows that zh>z~2>...>~r,., hence (2.5).
4 386 W. GAUTSCHI: Families We recall that the condition number in (2.2) is invariant with respect to scaling of the independent variable. Any transformation z=coz*, where co~=0 is arbitrary complex, leaves the condition of the roots unchanged. As a consequence, the assumption ]~,] = 1 made in (2.3) does not restrict generality, and we are free to subject the roots to a rigid rotation on the unit circle. Taking advantage of this last remark, we may bring the equation p~ (z)=0, which generally has complex coefficients, into a form with real coefficients. It suffices to rotate the roots into a position symmetric with respect to the real axis, i. e., to put Then 2 ~ e i(n-1)a/2 Z*. (2.8) p~ (e ic~-1~/2 Z*) "-- e i"(n-ll~'/z p,* (z*), where p~* (u)= /~ [u-~* (col,. x i(2--1--n--l~a ~* ta)=e T/, ~=1 and ~x*=~*.+l-~, 2=1,2... n, implying that all coefficients of p~* indeed are real. 3. Computation of the coefficients of p~* and the condition numbers cond ~. In analogy to (2.3), we write (3.1) p~*(u)= /7 [u--~*(a)]= 21 (--1)~*u ~-z, ~ro*=l, ).=1 ~0 where all crx* are real. Observing from (2.8) that p~ (z)--e incn-1)a/2 p~* (e -i(n-1)~/2 z), and using (3.1), we find i p~ (z)= 21 (--1) 4 e T M a~* z n-~. Hence, by comparison with (2.3), (3.2) az=e T M az*, 2=0, 1, 2... n.
5 of algebraic test equations For the following, it will be necessary to notation a2--o'4.., indeed, to define ~r~., by 387 introduce the more precise (3,3) /7 [z-r (a)] = Z (-1) 4 o'4,, z "-z, 1=I 4=0 /z----l, 2, 3... and similarly ~r*a,~ by (3.4) 1-1 [u-~*4,;, (~z)] -- X (-1) ~ a*4. u "-4, I=1 a~o where /z=l,2, 3... r a,f, (~z) = e i (4-1- e_~)=, I -- 1, 2... l.z. Then, as in (3.2), (3.5) O'2,tt = e T M a*2,/~ 9 Defining (3.6) a0,,=l, ~+x,~=0 for /z=0,1,2... one obtains from (3.3) the well-known recursion (3.7) a~,~+,=aa,~+ ~+i o'4_,,~, 2=1,2... /z+l, where ~+~=e ~=. Substituting (3.5) into (3.6), (3.7), and observing that all a*a,~ are real, one finds 0"*0,~ -" 1, O'*t,+l,~-- 0, o*z,u+~= cos (2a/2) a*z,u+ cos ((/z--2+ 1)a/2) a*z-l,~ ;z=0. 1, ,2... /z+l A simple induction argument will show that a*z,u--a*u_z., 2=0, 1,2... /.t. It suffices, therefore, to compute a*o,u-- 1, /z---0, 1... n-- 1,
6 388 W. GAUTSCHI: Families ~*x,~+l: cos (~.a/2) a*~,.+ cos ((#--Z + 1) a/2) G*~_,,.' 1 (5.8),l = 1, 2... ~ ' ~ = 0,1... n- 1. a~ =o",,* -1 ~-i,~,+1 7' **-1- (if /z is even) 1 / The numerator in the condition number (2.4), by (5.2), can now be obtained from ~-1 2 i1+ Ion,, +2 Z I~ a,,j, n even, -- 0"* "--" z t<,nl ~.=1 4=1 n Z [~z%,.[, n odd. In Figure 1 we show the largest condition number, cond r of a--2ztx/n, O_<x_<l, for n=10,20,40. log,o(cond) as function n =40 20 n - 20,, v, I,,, i I I I ~ ~"~'~'I, I ~ X Fig. 1. Condition number of the worst-conditioned root of p~ (z)=0, ~z=~.z~,xln, for 0<x<l, n- 10, 20, 40.
7 o/ algebraic test equations Equations with integer coefficients. The polynomials p~*, as constructed in Section 3, will not have integer coefficients, in general. It is desirable to find a subset of parameter values a, and a suitable constant multiple p~** of p~* such that all coefficients o-~:** of p~** become integer-valued when a is restricted to the subset in question. We can achieve this by letting a 1 (4.1) cos~-=l----, p>0. P Then all cosine factors in (3.8), hence all a*~,,, become polynomials in p-1 with f x f ~. integer coefficients, since cos(m2)=tm(1--~-), m=1,2,3... where T,~ is the Chebyshev polynomial of the first kind. Multiplying through by an appropriate power of p we obtain a polynomial p~** whose coefficients cry,,** are polynomials in p with integer coefficients, hence integer-valued if p is an integer. Selecting p sufficiently large we can get a arbitrarily small, and thus produce equations which are arbitrarily ill-conditioned. The only serious limitation of this procedure is due to the finiteness of the machine arithmetic in which the integer p and, above all, the resulting integer coefficients, are to be represented. A lower limit on p is imposed by the condition a<_2rc/n, which translates into (4.2) 1 p>_ sin 2 L 2n In Table 1 the results are summarized for 2_<n_<8. Only the coefficients a~:**, 2<_ [n/2], are listed, as the others can be obtained by symmetry, a2.,n** = -- 0"* :~ O" ** O" ** O" ** ** n 0,n l,n 2,n O'3,n* * 0"4, n 2 p s 2 (p) 3 1r 5 (P) 1# 2ps2 (P) s 4 (P) 2s 3 (P) s 4 (P) 5 # r 2 s 5 (p) 2s4 (p) ss (p) 6 p9 p4 s2 (p) S6 (p) ps 5 (p) s6 (p) 7 p12 p5 S7 (p) /)2 S6 (p) S7 (p) 8 p16 4p 9 S 2 (P) S 8 (p) 41# S 7 (p) S 8 (P) 2s2 (p) s4 (p) s5 (p) ~ (p) s5 (p) s7 (p) ~ (p) 4ps 2 (P) s7 (P) s 8 (P) s6 (P) 2ss (p) s7 (p) ~ (p) s-[(p) TABLE l. The coefficients of the test polynomial p=** (u)= j~ (--1) z cry,,*** u n-z. ~.~0
8 390 W. OAUTSCHI: Families The notations used in Table 1 are as follows: s2 (p) : 2 (p -- 1) sa (p)= (p--2) (3p--2) s4 (p)=p2--4p+2 s5 (p) = 5p 4-40p p2_ 64p + 16 s6 (p) = 3p ~- 32p p 2-64p + 16 s7 (p) = 7p 6-112p p 4-960ff + 880p 2-384p + 64 ss (p) = p6_ 20pS + 106p 4-224p p z- 96p + 16 s6 (p) = p2_ 8p + 4 ss (p) -- p4_ 16p p-'-- 32p + 8. Inequality (4.2) imposes the constraints (4.3) p>_l, p_>2, p_>4, p_>6, p~8, p_>ll, p_>14 for n = 2, , respectively. Thus. for example, the family of test equations of degree 5, according to Table 1 (n=5), is p6 z 5_ p2 ss (p) z4+ 2s4 (p) S5 (p) Z 3-2S4 (p) S5 (p) z 2-1- p2 S5 (p) Z- if= O, p > Numerical tests. Test equations can be put to many uses. One of the more important ones is the comparative analysis of different algorithms (and their computer implementations) for solving equations. Since we can vary the numerical condition of the roots, for each fixed degree, we are able to observe the performance of these algorithms on equations of fixed degree under increasing pressures of illconditioning. Our main interest, here, being in the accuracy of the algorithms, it is interesting to observe how closely the relative errors in the results are going to conform with what one ideally expects, namely that all these errors are of
9 oj algebraic test equations 391 the order of the machine precision multiplied by the respective condition numbers. If ~+0 denotes an exact root of the equation, "~ the approximation to returned by the computer algorithm, and e denotes the machine precision (i. e., ~=2 -t, where t is the number of binary digits in the mantissa of the floatingpoint word), then for a good algorithm the quantity (,5.1) /z (~)=E I~[ cond ought to be of the order of magnitude 1, or even smaller (considering that cond is a somewhat conservative measure of the condition of ~). The larger # (~), the poorer the performance of the algorithm (with regard to accuracy). If we are interested in the performance of the algorithm on an equation, rather than an isolated root, we can measure it by the average (5.2) z (r H k=l where n is the degree of the equation and ~k its n (simple) roots. By way of illustration, we compare two subroutines, ZRPOLY and ZPOLR, cut cf the IMSL library (International Mathematical Statistical Libraries, Version 7). The first implements the three-stage algorithm of Jenkins and Traub [1970], while the other is based on Laguerre's method, as implemented by Smith [1967]. We compiled both subroutines on the CDC 6500 computer, using the FTN compiler (CDC Fortran Extended Compiler), and ran them on the test equations p~*=0 of degrees n=5,10,20,40, with a=2zrx/n, x=.2, The coefficients a*z.~ were generated by (3.8) in double precision, and then rounded to single precision. Likewise, we used double precision to compute reference values for the roots ~z* (a). The results are summarized in Table 2 (2). The last two columns exhibit, for ZRPOLY and ZPOLR, respectively, the values observed for the average measure/za~ in (5.2). The two preceding columns contain the smallest and the largest condition number associated with the n roots in question. It is seen that the performance of both algorithms is quite good, in general, even on ill-conditioned equations. ZPOLR yields consistently smaller values of/z,~ (with one exception, for n-'5, x=.6). ZRPOLY seems to have some difficulties maintaining accuracy when x is near 1.0 and n is large (i. e., for relatively well-conditioned equations!). (2) Numbers in parentheses indicate decimal exponents.
10 392 W. GAUTSCHI: Families TABLE 2. Performance of ZRPOLY and ZPOLR on the test equation p~*=0, ~z=2rcx/n, x=.2(.2)1., n=5, 10,20,40. n x min cond max cond Na~ ZRPOLY Nay ZPOLR (3).184 (4) (--1) (2).960 (2) (--1) (1).135 (2) (i).229 (1).750 (2) (2) (6).408 (8) (3).564 (5) (--1) (2).761 (3) (1).176 (2).845 (2) (2).200 (1) (12).436 (17) (--1) (7).412 (11) (4).489 (7) (I).188 (4).687 (1) (--1).500 (--1).409 (5).100 (1) (25).973 (55).185 (--11).161 (--11) (13).451 (25) (--2) (7).599 (15).225 (1).707 (--1) (3).426 (8).200 (3).804 (--1) (--1).250 (--1).554 (I0).200 (1) Interestingly, both routines do not fall to pieces when confronted with an extremely ill-conditioned equation, such as the one for n=40, x=.2. All roots are obtained with relative errors ranging between 30 and 120%. This accounts for the very small values of /~a~ shown in this case. Similar results (3) are observed on equations with integer coefficients (those of Table 1), where n--2, , and where p is varied between the smallest 9 admissible integer (cf. (4.3)) and the largest possible integer subject to exact representation of the coefficients ~,n** in CDC 6500 floating-point arithmetic. (a) One of our tests runs (for n=3, p=8xl07) uncovered an implementation error in one of the auxiliary routines called by ZRPOLY.
11 o] algebraic test equations A second family of test equations. We now briefly describe our second family of test equations, viz., (6.1) q~ (z) =0, q~ (z)= /7 [z--lz (a)] =z~--vl z~-l+...+( - 1)" v,, ~.~1 ~(a)=a -~, l<a<co. The equations (6.1) are well-conditioned, when a is large, and become progressively more ill-conditioned as a approaches 1. The condition number of g~, using Gautschi [1973, Thm. 3.1], is easily found to be (6.2) where 27"g+v_l Tg+n_v -- o~-v(v-1)/2 cond{~=, v=l,2...,n, ~-v-1 ~-n-v zr0-*-"l, ~zu // (l_a-z), /x=l,2... n. It follows that while Furthermore, cond fl---~ 1, condf~--~2 (v>l), as a--~ oo, cond~--~ as as1. ~-4 v--1 7~+tn-v cond~<2, v=l,2... n. ~'-v-i ~-n-v The bounds on the right are symmetric, i. e., invariant under the substitution v--~ n+ 1-v, and increase monotonically for v<n/2, attaining their maximum value at v = [n/2] + 1. The true values of the condition number behave similarly. We have, in particular, ["/-~] (1 +,a -z ~ i6.3) J~,<,~max cond ~ < ~=1/7 \~-X ] 9 The largest condition number is shown in Figure 2 as a function of a, for n=10,20,40.
12 .594 W. GAUTSCHI: Families log.,(cond) t t a Fig. 2. Condition number of the worst-conditioned root of %(z)=o, for o~1, n= 10, 20, 40. The coefficients v~=va,n in (6.1) are easily generated by recursion, as in (3.6), (3.7), (6.4) ~-0,g=l, v~+l,u=o, /z-----o, 1,2... 7:a,g+l=vx,u+.ff~+l T~-1,~, 2=1,2... /z+i, /z=o, 1,2... n--1. If we let 1 ~=1+--, p>o, P and multiply through by appropriate powers of p-f 1, the new coefficients z*a,n become polynomials in p with integer coefficients, hence integer-valued, if p is an integer. More specifically, they assume the form
13 oj algebraic test equations (1-4-1) (n--~.) (n+l--l~ (6.5) ~*~,.=p ~ (p+ 1) 2 "r~..** (p), 2=0, 1,2... n, where v~,.** are polynomials satisfying zo,~** (p) = 1, v~..** (p)='c**._~..(p),,~=0, 1,2... n. The polynomials v~,.**(p). 2=1,2... In/2], for 2_<n_<_8, are Table 3. shown in TABLE 3. The coefficients of the test polynomial q~* (z)= j~ (--1).~p),(~+l)/2 (p.jrl)(n-2)(n*l-2)/2 T).,n** (P) zn-2,l.~0 n Tl,n** T2,n** T3,n** T4,n** 2 h (P) 3 t 3 (p) 4 t 2 (p) t 4 (p) t 3 (p) t 4 (p) 5 t 5 (p) t 4 (p) t 5 (p) 6 t 2 (p) fi (p) t~ (p) t 3 (p) t~ (p) t 6 (p) 7 6 (P) ta (P) t6 (P) t7 (P) 8 t 2 (p) t 4 (p) t 8 (p) t 4 (p) t 7 (p) t 8 (p) t 2 (P) t4 (P) t 5 (P) t 6 (P) t5 (P) t6 (P) h (P) h (P) t~ (p) t 6 (p) t 7 (p) t 8 (p) t5 (p) t6 (p) t7 (p) ts (p) The notations used in Table 5 are as follows: h (p) =2p+ 1 h (p)=sp2+sp+ 1 t4 (p)=2p2+ 2p-t - 1 (p)=5r lotr loy+ 5p+ 1 t6 (p)=p2 +p+ 1 t7 (p)= 7p6 + 21pS + 35p4 + 35p3 + 21pZ + 7p+ 1 t8 (p)=2pa +4p3 +6p2 +4p+ 1.
14 596 W. GauTscm: Families Thus, for example, the family of test equations of degree 5, according to Table 3 (n = 5), is (p-j- 1) 15 zs--p (pq- 1) l~ ts (p) Z4q'-p 3 (pq- 1) 6 t4 (p) ts (p) z 3- --ff (p+ 1) 3 t4 (p) ts (p) z2+p ~~ (p+ 1) t5 (p) z--p's=o, p--> Numerical tests. The same two subroutines, as in Section 5, were tested on the equations q~*=0 of degrees n=5, 10,20,40, for selected values of a between 1 and 5. The results are reported in Table 4, in an outlay similar to the one in Table 2. On the whole, the subroutines perform similarly as on the first test equations, ZPOLR generally yielding smaller values of /xav, but the differences are not as pronounced as before. In the case n=40, a=5, a series of low-order coefficients underflows, making a meaningful test impossible.
15 of algebraic test equations 397 TABLE 4. Performance of ZRPOLY and ZPOLR on the test equation q~*=0,,, =1.05 (var.) 5.0, n=5, 10,20,40. n ~ min cond max cond g~. ZRPOLY g~ ZPOLR (6).137 (7).256 (2) (5).947 (5).977 (--1).564 (--1) (3).326 (4) (2).329 (3) (--1) (2).491 (2) (--1) (1).527 (1) (10).231 (12).381 (1).450 (--1) (7).584 (9).148 (1) (4).363 (6) (3).333 (4) (2).113 (3) (1).549 (1) (15).759 (19).231 (--2) go (10).324 (14) (5).206 (8) (3).105 (5) (2).136 (3) (1).550 (1) (19).148 (29).133 (--5).758 (--2) (--1) t.1.i87 (11).889 (I8) (--1) (5).116 (9) (3).124 (5) (2).136 (3) (1).550 (1) (--6)
16 398 W. GAUTSCHI: Families REFERENCES YV'. GAUTSCHI: On the condition o/ algebraic equations, Numer. Math. 21, (1973), W. GAUTSCm: Questions o~ numerical condition related to polynomials, in: <<Recent Advances in Numerical Analysis >~ (1978), C. de Boor and G. H. Golub, eds., Academic Press, New York, R. T. GREGORY and D. L. KARNEY: A Collection o~ Matrices ]or Testing Computational. Algorithms (1969), Wiley-Interscience, New York. ['Corrected reprint, R. E. Krieger Publ. Co., Huntington, N. Y., 1978]. M. A. JENKINS and 1. F. TRAUB: A three-stage algorithm ]or real polynomials using quadratic iteration, SIAM J. Numer. Anal. 7, (1970), M. A. JENKINS and J. F. TRaun: Principles for testing polynomial zerofinding programs, ACM Trans. Math. Software 1, (1975), B. T. SMITH: ZERPOL, a zero finding algorithm /or polynomials using Laguerre's method, M. S. Thesis, Depertment of Computer Science, University of Toronto, (1967). ]. H. WXLKINSON: Rounding Errors in Algebraic Processes (1963), Prentice-Hall, Englewood Cliffs, N. J..
Part II NUMERICAL MATHEMATICS
Part II NUMERICAL MATHEMATICS BIT 31 (1991). 438-446. QUADRATURE FORMULAE ON HALF-INFINITE INTERVALS* WALTER GAUTSCHI Department of Computer Sciences, Purdue University, West Lafayette, IN 47907, USA Abstract.
More informationNumerische Mathematik
Numer. Math. 52, 241-250 (1988) Numerische Mathematik 9 Springer-Verlag 1988 Lower Bounds for the Condition Number of Vandermonde Matrices* Walter Gautschi 1,** and Gabriele Inglese e 1 Department of Computer
More informationETNA Kent State University
Electronic Transactions on Numerical Analysis. Volume 2, pp. 66-87, 2. Copyright 2,. ISSN 68-963. ETNA NUMERICAL CONDITION OF POLYNOMIALS IN DIFFERENT FORMS HONG ZHANG Abstract. The zeros of high-degree
More informationLet x be an approximate solution for Ax = b, e.g., obtained by Gaussian elimination. Let x denote the exact solution. Call. r := b A x.
ESTIMATION OF ERROR Let x be an approximate solution for Ax = b, e.g., obtained by Gaussian elimination. Let x denote the exact solution. Call the residual for x. Then r := b A x r = b A x = Ax A x = A
More information27r---./p. IR.(f)l<=,, l(o) max If(z)l. (1.2) o={z" z= 1/2(u + u-1), u p ei, O<- O <-- 2.a }
SIAM J. NUMER. ANAL. Vol. 27, No. 1, pp. 219-224, February 1990 1990 Society for Industrial and Applied Mathematics 015 A NOTE ON THE CONTOUR INTEGRAL REPRESENTATION OF THE REMAINDER TERM FOR A GAUSS-CHEBYSHEV
More informationModule 10.1: nag polynom eqn Roots of Polynomials. Contents
Nonlinear Equations Module Contents Module 10.1: nag polynom eqn Roots of Polynomials nag polynom eqn provides a procedure for computing the roots of a polynomial with real or complex coefficients. Contents
More informationCompute the behavior of reality even if it is impossible to observe the processes (for example a black hole in astrophysics).
1 Introduction Read sections 1.1, 1.2.1 1.2.4, 1.2.6, 1.3.8, 1.3.9, 1.4. Review questions 1.1 1.6, 1.12 1.21, 1.37. The subject of Scientific Computing is to simulate the reality. Simulation is the representation
More informationHOW AND HOW NOT TO CHECK GAUSSIAN QUADRATURE FORMULAE *
BIT 23 (1983), 209--216 HOW AND HOW NOT TO CHECK GAUSSIAN QUADRATURE FORMULAE * WALTER GAUTSCHI Department of Computer Sciences, Purdue University, West Lafayette, Indiana 47907, U.S.A. Abstract. We discuss
More informationChapter 1: Introduction and mathematical preliminaries
Chapter 1: Introduction and mathematical preliminaries Evy Kersalé September 26, 2011 Motivation Most of the mathematical problems you have encountered so far can be solved analytically. However, in real-life,
More informationNotes on floating point number, numerical computations and pitfalls
Notes on floating point number, numerical computations and pitfalls November 6, 212 1 Floating point numbers An n-digit floating point number in base β has the form x = ±(.d 1 d 2 d n ) β β e where.d 1
More informationFLOATING POINT ARITHMETHIC - ERROR ANALYSIS
FLOATING POINT ARITHMETHIC - ERROR ANALYSIS Brief review of floating point arithmetic Model of floating point arithmetic Notation, backward and forward errors Roundoff errors and floating-point arithmetic
More informationFloating-point Computation
Chapter 2 Floating-point Computation 21 Positional Number System An integer N in a number system of base (or radix) β may be written as N = a n β n + a n 1 β n 1 + + a 1 β + a 0 = P n (β) where a i are
More informationSOLVING SOLVABLE QUINTICS. D. S. Dummit
D. S. Dummit Abstract. Let f(x) = x 5 + px 3 + qx + rx + s be an irreducible polynomial of degree 5 with rational coefficients. An explicit resolvent sextic is constructed which has a rational root if
More informationJim Lambers MAT 610 Summer Session Lecture 2 Notes
Jim Lambers MAT 610 Summer Session 2009-10 Lecture 2 Notes These notes correspond to Sections 2.2-2.4 in the text. Vector Norms Given vectors x and y of length one, which are simply scalars x and y, the
More informationMathematical preliminaries and error analysis
Mathematical preliminaries and error analysis Tsung-Ming Huang Department of Mathematics National Taiwan Normal University, Taiwan September 12, 2015 Outline 1 Round-off errors and computer arithmetic
More informationFLOATING POINT ARITHMETHIC - ERROR ANALYSIS
FLOATING POINT ARITHMETHIC - ERROR ANALYSIS Brief review of floating point arithmetic Model of floating point arithmetic Notation, backward and forward errors 3-1 Roundoff errors and floating-point arithmetic
More informationUNCLASSIFIED Reptoduced. if ike ARMED SERVICES TECHNICAL INFORMATION AGENCY ARLINGTON HALL STATION ARLINGTON 12, VIRGINIA UNCLASSIFIED
. UNCLASSIFIED. 273207 Reptoduced if ike ARMED SERVICES TECHNICAL INFORMATION AGENCY ARLINGTON HALL STATION ARLINGTON 12, VIRGINIA UNCLASSIFIED NOTICE: When government or other drawings, specifications
More informationChapter 1 Error Analysis
Chapter 1 Error Analysis Several sources of errors are important for numerical data processing: Experimental uncertainty: Input data from an experiment have a limited precision. Instead of the vector of
More informationNumber Systems III MA1S1. Tristan McLoughlin. December 4, 2013
Number Systems III MA1S1 Tristan McLoughlin December 4, 2013 http://en.wikipedia.org/wiki/binary numeral system http://accu.org/index.php/articles/1558 http://www.binaryconvert.com http://en.wikipedia.org/wiki/ascii
More informationStandard forms for writing numbers
Standard forms for writing numbers In order to relate the abstract mathematical descriptions of familiar number systems to the everyday descriptions of numbers by decimal expansions and similar means,
More informationTHE NUMERICAL EVALUATION OF THE MAXIMUM-LIKELIHOOD ESTIMATE OF A SUBSET OF MIXTURE PROPORTIONS*
SIAM J APPL MATH Vol 35, No 3, November 1978 1978 Society for Industrial and Applied Mathematics 0036-1399/78/3503-0002 $0100/0 THE NUMERICAL EVALUATION OF THE MAXIMUM-LIKELIHOOD ESTIMATE OF A SUBSET OF
More informationPrentice Hall Algebra Correlated to: South Dakota Mathematics Standards, (Grades 9-12)
High School Algebra Indicator 1: Use procedures to transform algebraic expressions. 9-12.A.1.1. (Comprehension)Write equivalent forms of algebraic expressions using properties of the set of real numbers.
More informationCHAPTER 8: EXPLORING R
CHAPTER 8: EXPLORING R LECTURE NOTES FOR MATH 378 (CSUSM, SPRING 2009). WAYNE AITKEN In the previous chapter we discussed the need for a complete ordered field. The field Q is not complete, so we constructed
More informationComputer Arithmetic. MATH 375 Numerical Analysis. J. Robert Buchanan. Fall Department of Mathematics. J. Robert Buchanan Computer Arithmetic
Computer Arithmetic MATH 375 Numerical Analysis J. Robert Buchanan Department of Mathematics Fall 2013 Machine Numbers When performing arithmetic on a computer (laptop, desktop, mainframe, cell phone,
More informationHMMT February 2018 February 10, 2018
HMMT February 018 February 10, 018 Algebra and Number Theory 1. For some real number c, the graphs of the equation y = x 0 + x + 18 and the line y = x + c intersect at exactly one point. What is c? 18
More information27 th Annual ARML Scrimmage
27 th Annual ARML Scrimmage Featuring: Howard County ARML Team (host) Baltimore County ARML Team ARML Team Alumni Citizens By Raymond Cheong May 23, 2012 Reservoir HS Individual Round (10 min. per pair
More informationMATH 117 LECTURE NOTES
MATH 117 LECTURE NOTES XIN ZHOU Abstract. This is the set of lecture notes for Math 117 during Fall quarter of 2017 at UC Santa Barbara. The lectures follow closely the textbook [1]. Contents 1. The set
More informationFloating Point Number Systems. Simon Fraser University Surrey Campus MACM 316 Spring 2005 Instructor: Ha Le
Floating Point Number Systems Simon Fraser University Surrey Campus MACM 316 Spring 2005 Instructor: Ha Le 1 Overview Real number system Examples Absolute and relative errors Floating point numbers Roundoff
More informationw = X ^ = ^ ^, (i*i < ix
A SUMMATION FORMULA FOR POWER SERIES USING EULERIAN FRACTIONS* Xinghua Wang Math. Department, Zhejiang University, Hangzhou 310028, China Leetsch C. Hsu Math. Institute, Dalian University of Technology,
More informationOn the Condition of Algebraic Equations*
Numer. Math. 2t, 405--424 (t973) 9 by Springer-Veflag t973 On the Condition of Algebraic Equations* Walter Gautschi** Received March 12, t973 Summary. Given an algebraic equation, in which the polynomial
More informationRatio-Like and Recurrence Relation Tests for Convergence of Series
J. Inst. Maths Applies (1980) 25,349-359 Ratio-Like and Recurrence Relation Tests for Convergence of Series Y. F. CHANG AND G. CORLISS Department of Computer Science, University of Nebraska, Lincoln, Nebraska
More informationMath 128A: Homework 2 Solutions
Math 128A: Homework 2 Solutions Due: June 28 1. In problems where high precision is not needed, the IEEE standard provides a specification for single precision numbers, which occupy 32 bits of storage.
More informationMATHEMATICS OF COMPUTATION, VOLUME 26, NUMBER 118, APRIL An Algorithm for Computing Logarithms. and Arctangents. By B. C.
MATHEMATICS OF COMPUTATION, VOLUME 26, NUMBER 118, APRIL 1972 An Algorithm for Computing Logarithms and Arctangents By B. C. Carlson Abstract. An iterative algorithm with fast convergence can be used to
More informationFunctions and their Graphs
Chapter One Due Monday, December 12 Functions and their Graphs Functions Domain and Range Composition and Inverses Calculator Input and Output Transformations Quadratics Functions A function yields a specific
More informationLecture 7. Floating point arithmetic and stability
Lecture 7 Floating point arithmetic and stability 2.5 Machine representation of numbers Scientific notation: 23 }{{} }{{} } 3.14159265 {{} }{{} 10 sign mantissa base exponent (significand) s m β e A floating
More informationGeneralized eigenspaces
Generalized eigenspaces November 30, 2012 Contents 1 Introduction 1 2 Polynomials 2 3 Calculating the characteristic polynomial 5 4 Projections 7 5 Generalized eigenvalues 10 6 Eigenpolynomials 15 1 Introduction
More informationAlgebra Topic Alignment
Preliminary Topics Absolute Value 9N2 Compare, order and determine equivalent forms for rational and irrational numbers. Factoring Numbers 9N4 Demonstrate fluency in computations using real numbers. Fractions
More informationNumerische MathemalJk
Numer. Math. 44, 53-6 (1984) Numerische MathemalJk 9 Springer-Verlag 1984 Discrete Approximations to Spherically Symmetric Distributions* Dedicated to Fritz Bauer on the occasion of his 6th birthday Walter
More informationPre-Calculus MATH 119 Fall Section 1.1. Section objectives. Section 1.3. Section objectives. Section A.10. Section objectives
Pre-Calculus MATH 119 Fall 2013 Learning Objectives Section 1.1 1. Use the Distance Formula 2. Use the Midpoint Formula 4. Graph Equations Using a Graphing Utility 5. Use a Graphing Utility to Create Tables
More information27 Wyner Math 2 Spring 2019
27 Wyner Math 2 Spring 2019 CHAPTER SIX: POLYNOMIALS Review January 25 Test February 8 Thorough understanding and fluency of the concepts and methods in this chapter is a cornerstone to success in the
More informationProof of a conjecture by Ðoković on the Poincaré series of the invariants of a binary form
Proof of a conjecture by Ðoković on the Poincaré series of the invariants of a binary form A. Blokhuis, A. E. Brouwer, T. Szőnyi Abstract Ðoković [3] gave an algorithm for the computation of the Poincaré
More informationA Method for Reducing Ill-Conditioning of Polynomial Root Finding Using a Change of Basis
Portland State University PDXScholar University Honors Theses University Honors College 2014 A Method for Reducing Ill-Conditioning of Polynomial Root Finding Using a Change of Basis Edison Tsai Portland
More informationACM 106a: Lecture 1 Agenda
1 ACM 106a: Lecture 1 Agenda Introduction to numerical linear algebra Common problems First examples Inexact computation What is this course about? 2 Typical numerical linear algebra problems Systems of
More informationPre-calculus 12 Curriculum Outcomes Framework (110 hours)
Curriculum Outcomes Framework (110 hours) Trigonometry (T) (35 40 hours) General Curriculum Outcome: Students will be expected to develop trigonometric reasoning. T01 Students will be expected to T01.01
More informationMathematics for Engineers. Numerical mathematics
Mathematics for Engineers Numerical mathematics Integers Determine the largest representable integer with the intmax command. intmax ans = int32 2147483647 2147483647+1 ans = 2.1475e+09 Remark The set
More informationarxiv:math/ v1 [math.na] 12 Jul 2004
arxiv:math/0407177v1 [math.na] 12 Jul 2004 On improving the accuracy of Horner s and Goertzel s algorithms Alica Smoktunowicz and Iwona Wróbel Faculty of Mathematics and Information Science, Warsaw University
More informationIntegrated Math III. IM3.1.2 Use a graph to find the solution set of a pair of linear inequalities in two variables.
Standard 1: Algebra and Functions Students solve inequalities, quadratic equations, and systems of equations. They graph polynomial, rational, algebraic, and piece-wise defined functions. They graph and
More informationInterval solutions for interval algebraic equations
Mathematics and Computers in Simulation 66 (2004) 207 217 Interval solutions for interval algebraic equations B.T. Polyak, S.A. Nazin Institute of Control Sciences, Russian Academy of Sciences, 65 Profsoyuznaya
More informationErrors. Intensive Computation. Annalisa Massini 2017/2018
Errors Intensive Computation Annalisa Massini 2017/2018 Intensive Computation - 2017/2018 2 References Scientific Computing: An Introductory Survey - Chapter 1 M.T. Heath http://heath.cs.illinois.edu/scicomp/notes/index.html
More informationAn Alternative Proof of the Greville Formula
JOURNAL OF OPTIMIZATION THEORY AND APPLICATIONS: Vol. 94, No. 1, pp. 23-28, JULY 1997 An Alternative Proof of the Greville Formula F. E. UDWADIA1 AND R. E. KALABA2 Abstract. A simple proof of the Greville
More informationIntroduction and mathematical preliminaries
Chapter Introduction and mathematical preliminaries Contents. Motivation..................................2 Finite-digit arithmetic.......................... 2.3 Errors in numerical calculations.....................
More informationThe numerical stability of barycentric Lagrange interpolation
IMA Journal of Numerical Analysis (2004) 24, 547 556 The numerical stability of barycentric Lagrange interpolation NICHOLAS J. HIGHAM Department of Mathematics, University of Manchester, Manchester M13
More informationError Analysis of the Algorithm for Shifting the Zeros of a Polynomial by Synthetic Division
MATHEMATICS OF COMPUTATION, VOLUME 25, NUMBER 113, JANUARY, 1971 Error Analysis of the Algorithm for Shifting the Zeros of a Polynomial by Synthetic Division By G. W. Stewart III* Abstract. An analysis
More informationA NOTE ON THE LOCATION OF CRITICAL POINTS OF POLYNOMIALS
PROCEEDINGS OF THE AMERICAN MATHEMATICAL Volume 27, No. 2, February 1971 SOCIETY A NOTE ON THE LOCATION OF CRITICAL POINTS OF POLYNOMIALS E. B. SAFF AND J. B. TWOMEY Abstract. Let(P(a, 3) denote the set
More informationAccelerating the Numerical Computation of Positive Roots of Polynomials using Improved Bounds
Accelerating the Numerical Computation of Positive Roots of Polynomials using Improved Bounds Kinji Kimura 1, Takuto Akiyama, Hiroyuki Ishigami 3, Masami Takata 4, and Yoshimasa Nakamura 5 1,,3,5 Graduate
More informationMA257: INTRODUCTION TO NUMBER THEORY LECTURE NOTES
MA257: INTRODUCTION TO NUMBER THEORY LECTURE NOTES 2018 57 5. p-adic Numbers 5.1. Motivating examples. We all know that 2 is irrational, so that 2 is not a square in the rational field Q, but that we can
More informationESO 208A: Computational Methods in Engineering. Saumyen Guha
ESO 208A: Computational Methods in Engineering Introduction, Error Analysis Saumyen Guha Department of Civil Engineering IIT Kanpur What is Computational Methods or Numerical Methods in Engineering? Formulation
More informationERROR ANALYSIS FOR THE COMPUTATION OF ZEROS OF REGULAR COULOMB WAVE FUNCTION AND ITS FIRST DERIVATIVE
MATHEMATICS OF COMPUTATION Volume 70, Number 35, Pages 1195 104 S 005-5718(00)0141- Article electronically published on March 4, 000 ERROR ANALYSIS FOR THE COMPUTATION OF ZEROS OF REGULAR COULOMB WAVE
More informationCME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 5. Ax = b.
CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 5 GENE H GOLUB Suppose we want to solve We actually have an approximation ξ such that 1 Perturbation Theory Ax = b x = ξ + e The question is, how
More informationSolutions to Practice Final
s to Practice Final 1. (a) What is φ(0 100 ) where φ is Euler s φ-function? (b) Find an integer x such that 140x 1 (mod 01). Hint: gcd(140, 01) = 7. (a) φ(0 100 ) = φ(4 100 5 100 ) = φ( 00 5 100 ) = (
More informationFor δa E, this motivates the definition of the Bauer-Skeel condition number ([2], [3], [14], [15])
LAA 278, pp.2-32, 998 STRUCTURED PERTURBATIONS AND SYMMETRIC MATRICES SIEGFRIED M. RUMP Abstract. For a given n by n matrix the ratio between the componentwise distance to the nearest singular matrix and
More informationNumerical Algorithms. IE 496 Lecture 20
Numerical Algorithms IE 496 Lecture 20 Reading for This Lecture Primary Miller and Boxer, Pages 124-128 Forsythe and Mohler, Sections 1 and 2 Numerical Algorithms Numerical Analysis So far, we have looked
More informationNotes for Chapter 1 of. Scientific Computing with Case Studies
Notes for Chapter 1 of Scientific Computing with Case Studies Dianne P. O Leary SIAM Press, 2008 Mathematical modeling Computer arithmetic Errors 1999-2008 Dianne P. O'Leary 1 Arithmetic and Error What
More informationChapter 1 Computer Arithmetic
Numerical Analysis (Math 9372) 2017-2016 Chapter 1 Computer Arithmetic 1.1 Introduction Numerical analysis is a way to solve mathematical problems by special procedures which use arithmetic operations
More informationCorrections and emendations for Elementary Numerical Analysis, 3rd ed. by S. D. Conte and Carl de Boor 20feb18
Corrections and emendations for Elementary Numerical Analysis, 3rd ed. by S. D. Conte and Carl de Boor 20feb18 Each item in this list of corrections and emendations is in the form a/b/c: A = B [C] to indicate
More informationLebesgue Measure on R n
CHAPTER 2 Lebesgue Measure on R n Our goal is to construct a notion of the volume, or Lebesgue measure, of rather general subsets of R n that reduces to the usual volume of elementary geometrical sets
More informationMath Requirements for applicants by Innopolis University
Math Requirements for applicants by Innopolis University Contents 1: Algebra... 2 1.1 Numbers, roots and exponents... 2 1.2 Basics of trigonometry... 2 1.3 Logarithms... 2 1.4 Transformations of expressions...
More informationPre-Calculus and Trigonometry Capacity Matrix
Review Polynomials A1.1.4 A1.2.5 Add, subtract, multiply and simplify polynomials and rational expressions Solve polynomial equations and equations involving rational expressions Review Chapter 1 and their
More informationA Note on the Recursive Calculation of Incomplete Gamma Functions
A Note on the Recursive Calculation of Incomplete Gamma Functions WALTER GAUTSCHI Purdue University It is known that the recurrence relation for incomplete gamma functions a n, x, 0 a 1, n 0,1,2,..., when
More informationLecture 19 IIR Filters
Lecture 19 IIR Filters Fundamentals of Digital Signal Processing Spring, 2012 Wei-Ta Chu 2012/5/10 1 General IIR Difference Equation IIR system: infinite-impulse response system The most general class
More informationLinear Fractionally Damped Oscillator. Mark Naber a) Department of Mathematics. Monroe County Community College. Monroe, Michigan,
Linear Fractionally Damped Oscillator Mark Naber a) Department of Mathematics Monroe County Community College Monroe, Michigan, 48161-9746 In this paper the linearly damped oscillator equation is considered
More informationIntroduction CSE 541
Introduction CSE 541 1 Numerical methods Solving scientific/engineering problems using computers. Root finding, Chapter 3 Polynomial Interpolation, Chapter 4 Differentiation, Chapter 4 Integration, Chapters
More informationCS 450 Numerical Analysis. Chapter 8: Numerical Integration and Differentiation
Lecture slides based on the textbook Scientific Computing: An Introductory Survey by Michael T. Heath, copyright c 2018 by the Society for Industrial and Applied Mathematics. http://www.siam.org/books/cl80
More informationScattering Parameters
Berkeley Scattering Parameters Prof. Ali M. Niknejad U.C. Berkeley Copyright c 2016 by Ali M. Niknejad September 7, 2017 1 / 57 Scattering Parameters 2 / 57 Scattering Matrix Voltages and currents are
More informationChapter 6. Algebraic eigenvalue problems Introduction Introduction 113. Das also war des Pudels Kern!
6.0. Introduction 113 Chapter 6 Algebraic eigenvalue problems Das also war des Pudels Kern! GOETHE. 6.0. Introduction Determination of eigenvalues and eigenvectors of matrices is one of the most important
More informationAcademic Outcomes Mathematics
Academic Outcomes Mathematics Mathematic Content Standards Overview: TK/ Kindergarten Counting and Cardinality: Know number names and the count sequence. Count to tell the number of objects. Compare numbers.
More informationINTERVAL MATHEMATICS TECHNIQUES FOR CONTROL THEORY COMPUTATIONS
INTERVAL MATHEMATICS TECHNIQUES FOR CONTROL THEORY COMPUTATIONS R. Baker Kearfott Department of Mathematics University of Southwestern Louisiana Lafayette, Louisiana 70504-1010 1 Introduction and motivation
More informationElements of Floating-point Arithmetic
Elements of Floating-point Arithmetic Sanzheng Qiao Department of Computing and Software McMaster University July, 2012 Outline 1 Floating-point Numbers Representations IEEE Floating-point Standards Underflow
More informationIn Z: x + 3 = 2 3x = 2 x = 1 No solution In Q: 3x = 2 x 2 = 2. x = 2 No solution. In R: x 2 = 2 x = 0 x = ± 2 No solution Z Q.
THE UNIVERSITY OF NEW SOUTH WALES SCHOOL OF MATHEMATICS AND STATISTICS MATH 1141 HIGHER MATHEMATICS 1A ALGEBRA. Section 1: - Complex Numbers. 1. The Number Systems. Let us begin by trying to solve various
More informationKing s Year 12 Medium Term Plan for LC1- A-Level Mathematics
King s Year 12 Medium Term Plan for LC1- A-Level Mathematics Modules Algebra, Geometry and Calculus. Materials Text book: Mathematics for A-Level Hodder Education. needed Calculator. Progress objectives
More informationMCMULLEN S ROOT-FINDING ALGORITHM FOR CUBIC POLYNOMIALS
PROCEEDINGS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 130, Number 9, Pages 2583 2592 S 0002-9939(02)06659-5 Article electronically published on April 22, 2002 MCMULLEN S ROOT-FINDING ALGORITHM FOR CUBIC
More informationBP -HOMOLOGY AND AN IMPLICATION FOR SYMMETRIC POLYNOMIALS. 1. Introduction and results
BP -HOMOLOGY AND AN IMPLICATION FOR SYMMETRIC POLYNOMIALS DONALD M. DAVIS Abstract. We determine the BP -module structure, mod higher filtration, of the main part of the BP -homology of elementary 2- groups.
More informationNYS Algebra II and Trigonometry Suggested Sequence of Units (P.I's within each unit are NOT in any suggested order)
1 of 6 UNIT P.I. 1 - INTEGERS 1 A2.A.1 Solve absolute value equations and inequalities involving linear expressions in one variable 1 A2.A.4 * Solve quadratic inequalities in one and two variables, algebraically
More informationUtah Secondary Mathematics Core Curriculum Precalculus
A Correlation of Trigonometry Lial To the Utah Secondary Mathematics Core Curriculum Precalculus Resource Title: Trigonometry Publisher: Pearson Education Inc Publishing as Prentice Hall ISBN (10 or 13
More informationHonors Integrated Algebra/Geometry 3 Critical Content Mastery Objectives Students will:
Content Standard 1: Numbers, Number Sense, and Computation Place Value Fractions Comparing and Ordering Counting Facts Estimating and Estimation Strategies Determine an approximate value of radical and
More informationSolutions for Homework Assignment 2
Solutions for Homework Assignment 2 Problem 1. If a,b R, then a+b a + b. This fact is called the Triangle Inequality. By using the Triangle Inequality, prove that a b a b for all a,b R. Solution. To prove
More informationTHE TEACHER UNDERSTANDS THE REAL NUMBER SYSTEM AND ITS STRUCTURE, OPERATIONS, ALGORITHMS, AND REPRESENTATIONS
The real number SySTeM C O M P E T E N C Y 1 THE TEACHER UNDERSTANDS THE REAL NUMBER SYSTEM AND ITS STRUCTURE, OPERATIONS, ALGORITHMS, AND REPRESENTATIONS This competency section reviews some of the fundamental
More informationNumerical Methods. King Saud University
Numerical Methods King Saud University Aims In this lecture, we will... Introduce the topic of numerical methods Consider the Error analysis and sources of errors Introduction A numerical method which
More informationComplexity of Linear Programming. J.F. Traub Department of Computer Science Columbia University New York, N.Y H.
Complexity of Linear Programming J.F. Traub Department of Computer Science Columbia University New York, N.Y. 10027 H. Wozniakowski Institute of Informatics Department of Computer Science University of
More informationA Library of Functions
LibraryofFunctions.nb 1 A Library of Functions Any study of calculus must start with the study of functions. Functions are fundamental to mathematics. In its everyday use the word function conveys to us
More informationCHAPTER V Fermat s last theorem
5.1 Introduction. CHAPTER V Fermat s last theorem We discuss elementary methods approaches to Fermat s last theorem, in which the game is we do not use complex numbers. In this chapter we use methods available
More informationSECTION 1.4: FUNCTIONS. (See p.40 for definitions of relations and functions and the Technical Note in Notes 1.24.) ( ) = x 2.
SECTION 1.4: FUNCTIONS (Section 1.4: Functions) 1.18 (See p.40 for definitions of relations and functions and the Technical Note in Notes 1.24.) Warning: The word function has different meanings in mathematics
More informationNAG Library Routine Document C02AGF.1
NAG Library Routine Document Note: before using this routine, please read the Users Note for your implementation to check the interpretation of bold italicised terms and other implementation-dependent
More informationESTIMATION OF ERROR. r = b Abx a quantity called the residual for bx. Then
ESTIMATION OF ERROR Let bx denote an approximate solution for Ax = b; perhaps bx is obtained by Gaussian elimination. Let x denote the exact solution. Then introduce r = b Abx a quantity called the residual
More informationRoot Neighborhoods of a Polynomial
mathematics of computation volume 47. number 175 JULY WK6. PAÍIHS 2f>5-:73 Root Neighborhoods of a Polynomial By Ronald G. Mosier Abstract. The root neighborhoods of p(z), a polynomial over the complex
More informationA Note on Inverse Iteration
A Note on Inverse Iteration Klaus Neymeyr Universität Rostock, Fachbereich Mathematik, Universitätsplatz 1, 18051 Rostock, Germany; SUMMARY Inverse iteration, if applied to a symmetric positive definite
More informationPURE MATHEMATICS AM 27
AM Syllabus (014): Pure Mathematics AM SYLLABUS (014) PURE MATHEMATICS AM 7 SYLLABUS 1 AM Syllabus (014): Pure Mathematics Pure Mathematics AM 7 Syllabus (Available in September) Paper I(3hrs)+Paper II(3hrs)
More informationCore A-level mathematics reproduced from the QCA s Subject criteria for Mathematics document
Core A-level mathematics reproduced from the QCA s Subject criteria for Mathematics document Background knowledge: (a) The arithmetic of integers (including HCFs and LCMs), of fractions, and of real numbers.
More informationOffset Spheroidal Mirrors for Gaussian Beam Optics in ZEMAX
Offset Spheroidal Mirrors for Gaussian Beam Optics in ZEMAX Antony A. Stark and Urs Graf Smithsonian Astrophysical Observatory, University of Cologne aas@cfa.harvard.edu 1 October 2013 This memorandum
More informationA. Incorrect. Solve for a variable on the bottom by first moving it to the top. D. Incorrect. This answer has too many significant figures.
MCAT Physics - Problem Drill 03: Math for Physics Question No. 1 of 10 1. Solve the following equation for time, with the correct number of significant figures: Question #01 m 15.0 m 2.5 = s time (A) 0.17
More information